Re: [RFC] kern: simple futex for gnumach

2013-09-21 Thread Richard Braun
On Sun, Sep 22, 2013 at 01:01:31AM +0200, Samuel Thibault wrote: > Marin Ramesa, le Sat 21 Sep 2013 08:56:42 +0200, a écrit : > > + case FUTEX_WAIT: > > + return futex_wait(address); > > + break; > > > > + /* Save the value at the address. */ > > +

Re: [RFC] kern: simple futex for gnumach

2013-09-21 Thread Samuel Thibault
Marin Ramesa, le Sat 21 Sep 2013 08:56:42 +0200, a écrit : > + case FUTEX_WAIT: > + return futex_wait(address); > + break; > + /* Save the value at the address. */ > + int temp = *address; > + > + simple_lock(&futex_wait_lock); > + >

Re: [PATCH 2/4] coding style changes

2013-09-21 Thread Marin Ramesa
On 21.09.2013 16:13:24, Samuel Thibault wrote: > Marin Ramesa, le Tue 17 Sep 2013 19:35:03 +0200, a écrit : > > @@ -421,7 +421,7 @@ device_open(reply_port, reply_port_type, mode, > > name, device_p) > > ipc_port_t reply_port; > > mach_msg_type_name_t reply_port_type; > > dev_mode_t

Re: dbus startup problem when built with eglibc-2.17

2013-09-21 Thread Svante Signell
On Sat, 2013-09-21 at 15:59 +0200, Samuel Thibault wrote: > Svante Signell, le Mon 02 Sep 2013 22:35:51 +0200, a écrit : > > The solution is to patch dbus to cover also EPROTYYPE, > > if (*fd_p < 0 && (errno == EINVAL || errno == EPROTOTYPE)) > > if (retval < 0 && (errno == EINVAL || errno == E

Re: SOCK_CLOEXEC glibc patches (was: dbus startup problem when built with eglibc-2.17)

2013-09-21 Thread Samuel Thibault
Svante Signell, le Sat 21 Sep 2013 17:02:28 +0200, a écrit : > Strange, Pino just said on IRC that the patches were not accepted in > upstream libc, and the t/verify (whatever that is) might not even be > acceptable eglibc? t/verify is already as a patch debian's eglibc Samuel

Re: SOCK_CLOEXEC glibc patches (was: dbus startup problem when built with eglibc-2.17)

2013-09-21 Thread Svante Signell
On Sat, 2013-09-21 at 15:57 +0200, Samuel Thibault wrote: > Hello, > > Thomas Schwinge, le Mon 02 Sep 2013 23:56:08 +0200, a écrit : > > I have implemented SOCK_CLOEXEC for socket in TopGit branch > > t/socket_flags and for socketpair in t/socketpair_flags > > I have just added them to the debian

Re: Small cleanup of the device server routines

2013-09-21 Thread Samuel Thibault
Applied, except > [PATCH 2/4] coding style changes Thanks, Samuel

Re: [PATCH 2/4] coding style changes

2013-09-21 Thread Samuel Thibault
Marin Ramesa, le Tue 17 Sep 2013 19:35:03 +0200, a écrit : > @@ -421,7 +421,7 @@ device_open(reply_port, reply_port_type, mode, name, > device_p) > ipc_port_t reply_port; > mach_msg_type_name_t reply_port_type; > dev_mode_t mode; > - char * name; > + ch

Re: Small cleanup of the device pager

2013-09-21 Thread Samuel Thibault
Applied, except the cast coding style change. Thanks, Samuel

Re: Small cleanup of the device pager

2013-09-21 Thread Samuel Thibault
Marin Ramesa, le Tue 17 Sep 2013 04:44:45 +0200, a écrit : > The majority of casts are done without whitespace, so in patches 2 and > 4 I modified the rest to follow this rule. I don't think we want to follow that way. Samuel

Re: make /hurd/proc run as PID 3, do not use PID 0

2013-09-21 Thread Samuel Thibault
Justus Winter, le Mon 16 Sep 2013 12:20:24 +0200, a écrit : > Richard asked me to make /hurd/proc run with a PID other than 0 > because that causes various problems. I fully agree. Fortunately the > fix is straight forward. Applied, thanks. Samuel

Re: dbus startup problem when built with eglibc-2.17

2013-09-21 Thread Samuel Thibault
Svante Signell, le Mon 02 Sep 2013 22:35:51 +0200, a écrit : > The solution is to patch dbus to cover also EPROTYYPE, > if (*fd_p < 0 && (errno == EINVAL || errno == EPROTOTYPE)) > if (retval < 0 && (errno == EINVAL || errno == EPROTOTYPE)) > or to patch pf.c to return EINVAL instead of EPROTOT

Re: SOCK_CLOEXEC glibc patches (was: dbus startup problem when built with eglibc-2.17)

2013-09-21 Thread Samuel Thibault
Hello, Thomas Schwinge, le Mon 02 Sep 2013 23:56:08 +0200, a écrit : > I have implemented SOCK_CLOEXEC for socket in TopGit branch > t/socket_flags and for socketpair in t/socketpair_flags I have just added them to the debian glibc package. Samuel